Summary of Learning From Litigation: Graphs and Llms For Retrieval and Reasoning in Ediscovery, by Sounak Lahiri et al.
Learning from Litigation: Graphs and LLMs for Retrieval and Reasoning in eDiscovery
by Sounak Lahiri, Sumit Pai, Tim Weninger, Sanmitra Bhattacharya
First submitted to arxiv on: 29 May 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Information Retrieval (cs.IR)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper introduces DISCOvery Graph (DISCOG), a hybrid approach for electronic discovery that combines graph-based methods with Large Language Model (LLM) reasoning. Traditional approaches like BM25 and fine-tuned pre-trained models are limited by performance, computational, and interpretability challenges. In contrast, LLM-based methods prioritize interpretability but sacrifice performance and throughput. DISCOG uses a heterogeneous graph to generate embeddings and predict links, ranking the corpus for a given request. The approach outperforms baselines in F1-score, precision, and recall by an average of 12%, 3%, and 16%, respectively. In an enterprise context, DISCOG drastically reduces document review costs compared to manual processes and LLM-based classification methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine you have a huge pile of documents and need to find the important ones quickly. This paper shows how to use artificial intelligence (AI) to make that process faster and more accurate. The current methods used are good, but they can be slow and difficult to understand. The new approach, called DISCOvery Graph, uses two different AI techniques together to get better results. It’s like using a map to find the most important documents quickly and easily. This new method works really well and can save companies a lot of time and money by reducing the number of people needed to review the documents. |
Keywords
» Artificial intelligence » Classification » F1 score » Large language model » Precision » Recall
